The avian somatosensory system: connections of regions of body representation in the forebrain of the pigeon

Brain Research, 412 (1987) 205-223 Elsevier

205

BRE 12589

Research Reports

The avian somatosensory system" connections of regions of body representation in the forebrain of the pigeon J. Martin Wild Department of Anatomy, School of Medicine, University of Auckland, Auckland (New Zealand) (Accepted 21 October 1986)

Key words: Somatosensory; Hyperstriatum; Neostriatum; Thalamus; Wheatgerm agglutinin-horseradish peroxidase; Avian

In order to establish the basic connectivity of physiologically identified somatosensory regions of the thalamus and telencephalon in the pigeon, injections of wheatgerm agglutinin-horseradish peroxidase were made under electrophysiologicalcontrol and the projections were charted following conventional neurohistochemistry. The physiological recordings generally confirmed the findings of Delius and Bennetto (Brain Research, 37 (1972) 205-221) of somatosensory sites within the dorsal thalamus, anterior hyperstriatum and caudomedial neostriatum, and the anatomical results show that the thalamic cells of origin of the projections to the two telencephalic regions are largely separate: a rostral cell group comprising nucleus dorsalis intermedius ventralis anterior projects to the anterior hyperstriatum accessorium (HA), whilst a caudal cell group comprising caudal regions of nucleus dorsolateralis posterior (DLP) projects to the medial neostriatum intermedium and caudale (NI/NC). Caudal DLP is also the origin of a visual projection to NI/NC, and its terminal field also approximates that of the thalamic auditory nucleus ovoidalis. Since the anterior HA and NI/NC were here shown to be reciprocally connected, there is a possibility of multimodal input to both telencephalic regions. HA was also further defined as the origin of the basal branch of the septomesencephalic tract, and hence potentially provides an outlet for both telencephalic somatosensory regions. The results are discussed within a comparative context.

INTRODUCTION In mammals the s o m a t o s e n s o r y surface of the body is represented, like o t h e r sensory systems, in more than one m a j o r area of the cerebral neocortex, viz. the primary s o m a t o s e n s o r y area, traditionally known as SI, and the second s o m a t o s e n s o r y area, traditionally known as SII, both of which are now known to contain one or m o r e representations of the body depending on the species 13'37-39'59. But what of the organization of sensory systems in n o n - m a m m a lian vertebrates which do not have a cerebral neocortex of the m a m m a l i a n laminated type44'63? A similar b r o a d principle of 'multiple representation '36,58 may yet apply. In birds, for instance, there are at least two or three telencephalic areas for each of the visual, auditory and somatosensory systems 9'16,18,19,25,26, 28-30,43,46,53,64, but our knowledge of the anatomical

organization of the last, in particular, is far from complete. The beaks and oral cavity are r e p r e s e n t e d within a discrete basal telencephalic nucleus, the nucleus basalis, which receives its trigeminal input not from a thalamic nucleus but directly from the principal trigeminal nucleus 1°'23'82's3. The representation of the rest of the b o d y surface appears to be located in two other, quite separate telencephalic areas, namely the anterior h y p e r s t r i a t u m and the caudomedial neostriatum 18, but only in the case of the representation of the owl's claw in the anterior Wulst is the precise location of the thalamic input neurons known 5°. The present experiments were therefore carried out to d e t e r m i n e the basic pattern of connections of somatosensory regions of the forebrain. With respect to the thalamus particular attention was paid to the DLP, since it has been r e p o r t e d to receive a projection from the dorsal column nuclei 5'8° and spinal

Correspondence: J.M. Wild, Department of Anatomy, School of Medicine, University of Auckland, Auckland, New Zealand. 0006-8993/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)

206 cord 41 and from it can be recorded evoked potentials to natural and electrical stimulation of the body 18. DLP neurons can also be retrogradely labelled by the injection of H R P or fluorescent dyes into several regions of the telencephalon, including the hyperstriat u r n 6'62 and neostriatum 28'29'55. MATERIALS AND METHODS Each of 25 pigeons (Columba livia) was anesthetized either with a mixture of Ketamine (Ketalar, Parke Davis, 50 mg/kg) and Xylazine (Rompun, Bayer, 5 mg/ml) or urethane (1 ml of 15% ethyl carbamate per 100 ml body weight) and fixed in a Kopf stereotaxic apparatus. Monopolar tungsten microelectrodes (3-5 Mr2) were used (indifferent electrode attached to the head skin) to search the dorsal thalamus and/or anterior hyperstriatum and/or caudomedial neostriatum for evoked field potentials and single and/or multiple unit activity to natural stimulation of the body surface (feather tweaks, blowing, toe pinch, etc.) or to electrical stimulation of either the radial nerve, sciatic nerve, or both nerves alternately. The electrical stimulation (2-5 V, 0.1 ms pulses at 0.5 Hz) was applied by pairs of silver wire hook electrodes, cathode proximal, with the nerves covered in warm mineral oil. Potentials were usually amplified 1000 times, filtered with a bandpass of 10 Hz to 3 KHz, monitored on a dual beam oscilloscope and loudspeaker, and averaged with a Neurolog averager. Permanent records were obtained with an X - Y plotter. When maximal responses had been located within a particular region, the tungsten microelectrode was replaced with a glass micropipette (internal tip diameter 15-25 pro) filled with 3.5% wheatgerm agglutinate-horseradish peroxidase ( W G A HRP; Sigma) in 0.2 M Tris-HC1, 0.3 M NaCI buffer and the location re-identified by recording evoked potentials through the pipette. Constant positive current (2/zA) was then continuously applied to the electrode for 15-20 min, and the electrode left in place for a further 30 min prior to withdrawal. In two further cases a 'disconnection experiment' was carried out in order to test the presence of evoked potentials in the anterior hyperstriatum before and after severing telencephalic connections between the anterior hyperstriatum and caudomedial neostriatum. In one case a scalpel blade attached to

an electrode carrier was passed vertically through the neostriatum just anterior to the level of the anterior commissure, and in another the NI and overlying tissue was aspirated by gentle suction. The recording electrode was left in place during these procedures. After the recording of evoked potentials following the lesion, an iontophoretic injection of W G A - H R P was made at the recording site. Following a 24-48 h survival period the birds were deeply reanesthetized with an overdose of barbiturate and perfused through a common carotid artery with 250 ml saline containing 3% dextran and 2% sodium nitrite, followed by 500 ml of 3% glutaraldehyde in phosphate buffer (pH 7.4) containing 5% sucrose. The brains were blocked in the coronal plane 48 and immersed overnight in the fixative plus an additional 15% sucrose. Serial frozen sections, 50ktm, were collected in two series from the entire brain, reacted with tetramethyl benzidine 6°, mounted on subbed slides and one series was counterstained with thionin 2. A few sections through the center of the injection site were reacted with DAB in order to give a more accurate estimate of the center of the injection site, which appears smaller in DABreacted sections due to its relatively poor sensitivity. All sections were viewed with both light- and darkfield optics and the location of retrogradely labelled neurons and anterograde projections were drawn with the aid of a drawing tube. RESULTS

Evoked responses to tactile and electrical stimulation Thalamus. Within the dorsal thalamus, field potentials and/or single and multiple unit activity could be evoked within a region bounded by the following coordinates4S: A4.80-6.50, L1.5-2.3 and at depths of 7.20-7.90 from the dura. On electrical stimulation of peripheral nerves evoked field potentials 100-200 /~V in amplitude were obtained more anteriorly within this region and single or multiple spikes frequently accompanied the negative wave (Fig. 1A). More posteriorly within this region (A4.8-5.5) field potentials were much less readily evoked and were shallow, but single or multiple units were frequently encountered (Fig. 1B). Latencies either to the initial deflection of averaged field potentials or to single spikes were variable both within birds at different

207 the latency recorded by Delius and B e n n e t t o is, and 15 ms for leg stimulation.

~ B A5.00

soL_ O 5

Fig. 1. A, C, D: a single (upper trace) and an average of 64 (lower trace) field potentials evoked from the site marked with a star in the schematic hemisection at right. In each case the stimilu were 0.1-ms pulses at 0.5 Hz, 2-3 V applied either to the contralateral sciatic (A) or radial (C, D) nerve. Note the short latency in D. B: two traces of the same single unit recorded from cDLP. In the upper trace a 3-V, 0.1-ms pulse was applied to the contralateral sciatic nerve; in the lower trace an identical pulse was applied to the contralateral radial nerve. Calibrations are in ms and/~V. Coordinates for the hemisections (A5.75, etc.) are from Karten and Hodos48.

loci and between birds, and were not correlated with the type of anesthetic. To stimulation of the radial nerve, latencies ranged from 5 - 2 0 ms; to stimulation of the sciatic nerve 9 - 2 5 ms. M o d a l latencies were 9 ms for wing stimulation, which agrees perfectly with

Field potentials at any locus could frequently be driven from either side of the body, although contralateral stimulation generally p r o d u c e d larger and more consistent responses than ipsilateral stimulation, F u r t h e r m o r e , at some loci, field potentials or single unit activity could be r e c o r d e d following stimulation of both the radial and sciatic nerves (e.g. Fig. 1B), although field potential amplitudes were often greater to stimulation of one nerve than of the other. Natural stimulation confirmed these findings by showing that the receptive fields of single units were m o r e often than not rather large, extending for instance over a whole wing, shoulder, neck and head, or a leg and back. Some units, however, were exquisitely sensitive only to the deflection of a few feathers of the contralateral wing or to light brushing of the toes. Overall, a somatotopic organization within the dorsal thalamus was not obvious in this study although at m o r e anterior loci in some cases responses to leg stimulation were maximal at slightly d e e p e r coordinates than responses to wing stimulation. Telencephalon. Within the telencephalon, e v o k e d responses were r e c o r d e d in both the anterior hyperstriatum and caudomedial neostriatum. In the former, responses were most reliably e v o k e d in a restricted region of the H A , from A 1 2 . 5 - 1 4 . 0 , L 0 . 8 1.7 and 2 . 0 - 4 . 5 below the surface (Fig. 1C). Latencies again were variable but were typically in the order of 15-25 ms. In the latter, responses could be e v o k e d within a region defined by A5.75-7.5, L 0 . 8 - 3 . 0 and from 2 . 5 - 4 . 6 below the surface. The loci yielding the most vigorous multiunit responses to radial or sciatic stimulation were the deepest in this region, at 4.5 m m depth, immediately dorsal to the C I O (see Fig. 6 C - F ) . E v o k e d responses in this region were larger than in the H A , up to 300/~V, and their latencies could be considerably shorter, in some instances as short as 6 - 8 ms to stimulation of the radial nerve (Fig. 1D); in others as long as 16-18 ms. A s in the thalamus, some single units in both H A and caudomedial neostriatum could be driven by stimulation of the radial and sciatic nerves or from both sides of the body, but the receptive fields of others were very restricted, e.g. to a few feathers of the contralateral wing, and these were exquisitely sensi-

208 tive to the slightest deflection of a feather barbule. Also, within H A , some vertical electrode passes gave some suggestion of an ordered representation of the body. In these cases, as the electrode descended from the surface, units responsive to light tactile stimuli could be driven from successively more caudal, predominantly contralateral, parts of the body. Near the surface, units responsive to light stroking of the beak were encountered, then, a little lower, units responsive to minute deflections of feathers under the eye, then upper neck, lower neck, upper wing and so on down to the claws, where some units were responsive only to light touch, others only to squeezing the foot. At some but not at all sites within the caudomedial neostriatal region which were responsive to tactile stimulation of the body or to electrical stimulation of peripheral nerves, some units were also exquisitely sensitive to the lightest possible tactile stimulus which could manually be applied to the beaks, head and skull, beak bar, ear bar, stereotax or heavy table

f"-'-,-~L:IIS ~ HA

H

on which the stereotax was mounted. These units could not be driven, however, either by electrical stimulation of the beak surface via a pair of needle electrodes or of the surgically exposed mandibular branch of the trigeminal nerve via a pair of hook electrodes.

Anterograde and retrograde WGA-HRP labelling Thalamic injections. In 9 cases iontophoretic injections of W G A - H R P were made at various loci under electrophysiological control. Either one or the other of two major telencephalic zones of termination were consistently defined by these injections, one in the rostral H A and the other in the caudomedial neostriatum. With reference to the thalamic nuclei defined in the atlas of Karten and Hodos 48, injections centered on the most rostral and ventral parts of D L P and on ventral parts of D L A and D L M (n = 5; e.g. Figs. 2G and 4A), but in two cases encroaching on SPC dorsolaterally, produced terminal labelling within H A from approximately A13.0 to the ros-

A

~

'

~l~D~.lf:!

HA

OV

/4

Fig. 2. A rostrocaudal series (A-F, approximately 250/~m apart from the rostral pole of the brain) showing tracings of labelled fibers and terminals (fine dots) following an injection of WGA-HRP into the dorsal thalamus shown as the stippled area in G. H: darkfield photomicrograph showing terminal labelling in HA resulting from the same injection. Bar = 250 Bm.

209 tral pole of the brain (Fig. 2 A - F ) . Here, the density of extraperikaryal reaction product was not uniform but was particularly dense laterally and ventrolaterally within the I H A which borders the HIS (Fig. 2H). The density of labelling gradually diminished within H A toward the midline. In two cases the whole dorsoventral extent of H A was labelled, but in the other three the most dorsal part was either not labelled or only slightly labelled. These differences were presumably due to small differences in the placement or size of the injections. Labelled fibers linking the cells of origin of this projection with H A were observed to take the following course. The fibers left the injection site and gathered within the FPL. As they passed by the RSd, some terminations were given off there, and retrogradely labelled cells were also found in the most dorsolateral portion of this nucleus at about A7.25. Labelled fibers then entered the telencephalon on

HV

(

HV

the ventral aspect of FPL but continued rostrally through the mediobasal telencephalon. The rather fine fibers were not able to be visualized between approximately A9.0 and A l l . 0 , but in view of their destination and the point at which they were next observed they possibly take the course within the tractus fronto-thalamicus et tractus thalamo-frontalis described by Karten et al. 51 for efferent fibers of the dorsomedial thalamic nuclei. The labelled fibers were next observed gathering within the neostriatum on the dorsolateral aspect of the lobus parolfactorius from where they coursed dorsally and rostrally in a wide sheet through mid-portions of the neostriatum, hyperstriatum ventrale, and hyperstriatum dorsale. As they approached the rostral pole of the telencephalon they turned medially at right angles through HIS to terminate within H A (Fig. 2). Four injections centered on more caudal parts of DLP, from A5.5-4.5 (e.g. Figs. 3F and 4B), (in one

v

HV

~,==i~~,.:H V

.........

Fig. 3. A rostrocaudal series (A-E, approximately 500 #m apart) showing tracings of labelled fibers and terminals (fine dots) following an injection of WGA-HRP into the dorsal thalamus shown as the stippled area in F. G: a darkfield photomicrograph showing terminal labelling in the caudomedial neostriatum resulting from the same injection. Bar = 250/~m.

210 case encroaching on the medial pretectal area laterally and in two cases on the underlying SpM), produced terminal labelling in a medial neostriatal field which straddles the a n t e r o m e d i a l NC and c a u d o m e dial NI, hence a b b r e v i a t e d N I / N C (Fig. 3). H e r e the label e x t e n d e d as an arched b a n d from the region of the C I O at a p p r o x i m a t e l y A7.00 to the medial aspect of the caudal pole of the ectostriatum at approximately A9.00. In 3 cases this b a n d was frequently c o m p o s e d of two m o r e or less discontinuous patches (e.g. Fig. 3D), one centered medially on C I O (Fig. 3G), the other dorsolateral to C I O . In the case which had the smallest injection confined within the bounds of the caudal D L P nucleus, only the m o r e medial terminal patch was observed. The route of labelled fibers projecting from the injection site to NI/NC was as follows. Efferent fibers travelled rostrally through F P L and, like the projections from m o r e rostral thalamic regions, gave off some terminations within dorsal portions of R S d , but a little m o r e caudally, at A6.75. R e t r o g r a d e l y labelled cells were present within R S d also at this level, and at A7.00 where they were located just below the O M as it descends into the thalamus. Fibers then passed under O M and t u r n e d d o r s o m e d i a l l y to travel through the paleostriatum before commencing their terminations within N I / N C i m m e d i a t e l y dorsal to the L M D (Fig. 3G). In this study, injections which a n t e r o g r a d e l y labelled H A did not label N I / N C and vice versa, although in cases where SPC was involved in the injection, light terminal labelling was o b s e r v e d in the m o r e posterior Wulst in addition to that in anterior HA. Both sets of injections centered either on caudal D L P or rostral D L P , D L A / D L M p r o d u c e d retrograde labelling of neurons within the dorsal column and/or external cuneate nuclei (Fig. 5), corroborating the s o m a t o s e n s o r y input to these thalamic regionsS0, sl.

HA injections. Injections of W G A - H R P

Fig. 4. Lightfield photomicrographs of actual WGA-HRP injection sites in each of the 4 areas of interest in the present study. A: centered on rostroventral DLP and ventral

into the

DLM/DLA (nuclei according to Karten and Hodos% but see text and caption to Fig. 6. B: on cDLP. C: caudomedial neostriatum. D: HA. In C, the dark horizontal strip leaving the injection site to the left is artifactual staining along a tissue tear. The hemisphere is somewhat distorted in this case. All sites have visible electrode tracks running through or near them. Bars = 500 #m.

211

,4 •

U

,,a'

Fig. 5. A, left: schematic section of the caudal medulla showing the location of retrogradely labelled neurons (dots) in the gracile nucleus 81 following an injection of W G A - H R P into the contralateral dorsal thalamus such as is shown in Fig. 2G. A, right: a darkfield photomicrograph showing some of these labelled neurons. All are labelled neurons, in differential focus, except for the artifact identified with arrows. B, left: a schematic hemisection of the medulla rostral to the obex, showing the location of retrogradely labelled neurons (dots) in the external cuneate nucleus 8~, following an injection of W G A - H R P into the dorsal thalamus such as is shown in Fig. 3F. B, right: a lightfield photomicrograph showing some of these labelled neurons. Calibration bars = 50/~m.

rostral H A (n = 7) were either totally confined within its lateral boundary, or, if very ventral, encroached slightly on HIS (Figs. 6J and 4D). The location of retrogradely labelled neurons within the thalamus is shown for a typical case in Fig. 6 D - I . They can be seen to form a relatively compact nucleus (Fig. 6K) which extends over 1.25 m m rostrocaudally from a position immediately dorsal to FPL rostrally to a position dorsal to the nuclei rotundus and triangularis caudally. In the thalamus of the owl a somatosensory nucleus has been described in a similar location 5°, on which basis it was appropriately designated nucleus dorsalis intermedius ventralis anterior ( D I V A ) . This name will thus be adopted for the corresponding nucleus in the pigeon, which can be identified in normal, Nissl-

stained material (Fig. 7) as cytoarchitecturally distinct from D L A , D L M and rostral DLP, all of which lie immediately dorsal to D I V A throughout different regions of its rostrocaudal extent. Labelled D I V A neurons were uniformly small to medium in size (8-15/~m) and almost entirely ipsilateral except for an occasional contralaterally labelled cell located near the SPC border. Labelled neurons were also present in SPC bilaterally, with the contralateral neurons located slightly caudal to the ipsilateral neurons. A second group of neurons retrogradely labelled from H A injections was located ipsilaterally within a caudomedial neostriatal region which includes but is not confined to the zone of termination of the caudal D L P projection (Fig. 6 A - G ) .

212

~HV

HV

.:~.'

N

HV N .:,

HV N .i~-..

..

N

x"~"

DLM

H

F K

Fig. 6. A rostrocaudal series ( A - I , approximately 250#m apart) showing the location of retrogradely labelled neurons (dots) in the dorsal thalamic nucleus DIVA ( D - I ) and in the caudomedial neostriatum ( A - G ) following an injection of W G A - H R P into the rostroventral HA shown as the stippled area in J. K, L: darkfield photomicrographs showing retrogradely labelled neurons in DIVA and NI/NC, respectively, following the H A injection. Anterograde terminal labelling is also apparent in L. Bars = 200 #m.

213

FPL

Fig. 7. Photomicrographs and corresponding line drawings of normal, Nissl-stainedsections (50/~m) through a rostral (A) and a caudal

(B) level of the nucleus DIVA in the pigeon thalamus. Calibration bars = 500 ~m.

Within this caudomedial neostriatal zone was also evidence of terminal labelling scattered between the retrogradely labelled perikarya (Fig. 6L), indicating a reciprocal connection between H A and the caudomedial neostriatum. H A injections also labelled part of the major efferent pathway of the Wulst, namely the TSM. Labelled fibers were largely restricted to the lateral aspect of this tract as it leaves the base of the telencephalon to form the basal branch of TSM 4°. The major zone of termination of TSM in these cases was the SpM of the caudal thalamus. Only very sparse labelling of the tectum was observed, and no anterograde labelling was observed caudal to diencephalic levels in these cases. The H A projection to SpM was additionally indicated by the fact that the caudal DLP injections which encroached on SpM produced abundant retrograde labelling of neurons within the rostral HA. DLP did not receive a projection from HA.

Neostriatal injections. Injections of W G A - H R P located at somatosensory recording sites distributed at different loci within the caudomedial neostriatum (n = 7; e.g. Figs. 8L and 4C) produced retrograde labelling in 3 locations: (1) in the rostral HA, particularly its ventral regions (Fig. 8 A - G and M), thus confirming the anterograde results from H A injections; (2) in caudal DLP (cDLP; Fig. 8 H - K and N), particularly within the more ventral parts of this nucleus which is composed of a distinct cluster of larger cells (10-22/~m) lying immediately ventral to the lateral aspect of the HIP (= fasciculus retroflexus); and (3) in the Ov, the main thalamic auditory nucleus 12' 42,43, and in the TOv (Fig. 8H). Every neostriatal injection which labelled caudal DLP neurons also labelled neurons in Ov (n = 5), but two injections located more dorsally than the others within the neostriatum labelled Ov neurons without labelling any caudal DLP neurons. No labelled neurons were

214

M

tN,__ Fig. 8. A rostrocaudal series (A-K, approximately 250/am apart) showing the location of retrogradely labelled neurons (dots) in HA (A-G), in cDLP (H-K), and in Ov (H) following an injection of WGA-HRP into the caudomedial neostriatum shown as the stippled area in L. M, N: darkfield photomicrographs showing retrogradely labelled neurons in HA and cDLP, respectively, following the neostriatal injection. Light anterograde terminal labelling is also apparent in M. Bars = 50 Bm.

found in the pretectal area, SPC, SpM, D L A , D L M , D I V A or the most rostral parts of D L P following neostriatal injections at somatosensory recording sites. Anterograde labelling produced by neostriatal injections was also found in 3 locations. One was the

rostral H A (Fig. 8M), thus confirming the retrograde results from H A injections and the reciprocal nature of HA-caudomedial neostriatal connections. The most ventral neostriatal injections produced terminal labelling predominantly in the ventral part of H A , whereas more dorsal neostriatal injections produced

215 terminal labelling in more dorsal and medial H A regions. A second region labelled by neostriatal injections was the underlying PA 12,52 and a third was the neostriatum dorsal and caudal to the injection site, immediately ventral to the ventricle. This latter region corresponds in general location to that of a major song control nucleus of songbirds, the so-called hyperstriatum ventrale, pars caudale ( H V c ) 12'52'66. A fourth area of apparent terminations was in the dorsal regions of the nucleus reticularis superior, pars dorsalis, just rostrolateral to Ov. These presumed terminations took the form of large bulbous enlargements of labelled fibers travelling in the lateral forebrain bundle, but it was not possible to determine whether these were terminations of NI/NC efferent neurons, or of collaterals of retrogradely labelled Ov or cDLP neurons.

Disconnection experiments In these two cases electrical stimulation of the radial nerve produced typical evoked responses in HA. Immediately following the lesion in each case, these responses could still be evoked at the same prelesion site, although they were somewhat reduced in ampli-

tude, possibly due to some movement of the brain during lesioning. Following histochemical processing, the distribution of labelled neurons was seen to be identical to that following the other H A injections (see above). No labelled neurons were present in the neostriatum in the case in which this could be verified. DISCUSSION

Fig. 9 summarizes the connections of the avian somatosensory system as they now appear. The dorsal column-medial lemniscal system conveys somatosensory information from the periphery 81 to the dorsal thalamus 8°, as in mammals. Two different dorsal thalamic nuclei, both of which receive projections via the medial lemniscus and spinal cord, then project largely if not entirely exclusively to two widely separated but reciprocally interconnected telencephalic fields; a rostrally situated thalamic group to the anterior HA, and a caudally situated thalamic group to the caudomedial neostriatum.

DIVA-HA The rostral group comprises uniformly small to

NC

Cb

HV

HA

G~.CtCE BODY INPUT

~.. TO

k , . ~ x BEAK " INPUT Fig. 9. Some connections of the somatosensory system of the pigeon as they now appear. O F : origins; - - < : terminations; - - : somatosensory pathways; - - -: potential somatosensory pathways, and ascending auditory pathway from Ov (see text); t: actual trajectory unknown; *: see ref. 82 for further connections.

216 medium sized neurons of the nucleus D I V A which has afferent and efferent connections similar to its namesake in the owl 5°. Delius and Bennetto TMrecorded somatosensory responses in the dorsal thalamus of the pigeon in a region which undoubtedly included parts of D I V A , but they also found somatosensory responses throughout the overlying rostral DLP at A5.5-5.75. In the present study the rostral thalamic somatosensory area was found to extend more anteriorly than the region defined by Delius and Bennetto but responses were more ventrally restricted. One reason which could account for this latter difference is that Delius and Bennetto were primarily interested in the cutaneous representation and hence stimulated various (unspecified) skin nerves and the skin itself. Some of these nerves or areas might simply not have been represented within the areas studied in the present experiments. It should be noted, however, that Delius and Bennetto used concentric bipolar electrodes for the most part in contradistinction to the probably finer monopolar tungsten electrodes used in the present study. And the fact remains that the location of neurons retrogradely labelled from H A injections in the present study was largely coincident with the loci of recording sites which themselves were largely restricted to a narrow band immediately dorsal to FPL and the nuclei rotundus and triangularis, i.e. DIVA. Moreover, it is this area rather than more dorsal regions within rostral DLP, D L A or D L M which receives the greatest density of afferent terminations of the medial lemniscus s° and spinothalamic projections 41. These terminations are bilateral 8° although predominantly contralateral, but the projection from D I V A to H A appears to be almost totally ipsilateral. D I V A and HA, therefore, both contain a bilateral representation of the body but with a contralateral predominance (see also Delius and Bennetto18). But neither within D I V A nor within H A was there unequivocal evidence of a somatotopic organization as revealed by the present or previous TMtechniques, although some electrode tracks in H A in the present experiments did yield some evidence of an ordered body representation, with a dorsoventral sequence in H A corresponding to a rostrocaudal body topography. However, it does seem that most parts of the body are represented within D I V A and HA, includ-

ing the head and possibly the beaks as well. This last is interesting in view of the now well-known direct projections of the principal trigeminal nucleus - which is a major recipient of terminations of primary afferents from the beaks - - to the nucleus basalis of the telencephalon via the quintofrontal tract 23'82. Since this projection bypasses the thalamus, the question is how beak representation comes also to be included in HA, and presumably in D I V A 83. In the absence of connections between nucleus basalis and HA, the answer to this may well lie in the fact that a component of the descending projections of the gasserian ganglion terminates in the external cuneate nucleus 22, the major projection of which is not to the cerebellum but via the medial lemniscus to D I V A and other parts of the thalamus 8°. As defined, D I V A does not include neurons belonging to SPC, even though H A receives a bilateral projection from this nucleus. A projection to SPC from the medial lemniscus has not been noted 5,8°, but it is reported to receive a spinal projection 47. Delius and Bennetto TM found somatosensory responses in SPC but such were not found in the present study. The role of SPC in the somatosensory system remains to be clarified. It can be noted, however, that the number of neurons retrogradely labelled in SPC from H A injections in the present experiments was small compared with the number in D I V A , and yet SPC is quite a large nucleus with a considerable rostrocaudal extent and, as its name suggests, is densely packed with small neurons. It may well be functionally diverse. The terminal field of D I V A in H A as defined in the present experiments does not exactly coincide with the rostral somatosensory area defined purely on the basis of evoked potentials by Delius and Bennetto TM, although it is very close. These authors placed the site of maximal responses within deep regions of the hyperstriatum intercalatus superior (HIS), but it is clear from their description and their figure 1 that evoked potentials similar in amplitude and latency to those recorded in the present experiments, were also recorded in H A , particularly in the region close to the HIS border. This region was shown in the present study to receive the densest terminations of thalamic projections to the anterior hyperstriatum. It corresponds to the lamina more clearly distinguishable in the owl as the intercalated hyperstriatum accessori-

217 um 51 (IHA).

DLP-NI/NC The other dorsal thalamic cell group of the somatosensory system is composed of neurons within caudal regions of DLP (cDLP), particularly those which form a distinct cluster immediately ventral to the lateral aspect of the fasciculus retroflexus. In Nisslstained normal material this part of DLP is obviously different from more rostral regions of DLP 29 in that it contains many larger neurons. Closer inspection reveals, however, that many small neurons are also present, an observation confirmed in Fig. 8N which shows that caudal DLP neurons retrogradely labelled from NI/NC injections are remarkably diverse in size. Although bilateral responses were also recorded in this nucleus, they tended to be single units or very shallow field potentials, unlike the responses recorded in DIVA, and in fact the apparently important role of this nucleus in the somatosensory system was only fully appreciated in the present study when recordings and W G A - H R P injections were made in the terminal field of the caudal DLP projection within NI/NC. The somatosensory region of NI/NC was first discovered by Erulkar 26 and later confirmed by Delius and Bennetto 18. The neostriatal area defined by the latter authors, however, did not extend as far rostral or as deep as the region defined in the present experiments; in fact in the study by Delius and Bennetto ~8, evoked potentials were mainly recorded between A5.5 and A6.5 and were seen to disappear at 4.5 mm below the surface, whereas in the present study the most vigorous responses to radial or sciatic nerve stimulation occurred at precisely this depth, only more anteriorly. This ventral region lying within and near to CIO at A6.75-7.50, immediately dorsal to the LMD, coincides with the more medial of the two patches of dense terminal labelling produced by W G A - H R P injections centered on cDLP and was the only physiologically defined somatosensory locus within NI/NC at which W G A - H R P injections consistently and heavily retrogradely labelled caudal DLP neurons. Gamlin and Cohen 29 have shown that the rostral, small-celled part of DLP (rDLP) projects to the anterior NI medial to the ectostriatum, which, together

with the present results, provides evidence for the distinction of rDLP from the subjacent DIVA. These authors suggested, however, that the rDLP projection was a somatosensory one, but in the present study no somatosensory responses could be recorded from anterior regions of NI.

H A - N I / N C reciprocal connections Although the major somatosensory input to H A and NI/NC appears to be provided by D I V A and cDLP, respectively, the present novel findings of reciprocal ipsilateral connections between NI/NC and H A raise the possibility that the somatosensory input which reaches these areas could arise over two routes, one direct, the other indirect. For instance, the somatosensory input to H A could arrive directly from D I V A or indirectly via caudal DLP and NI/NC. The two disconnection experiments do not rule out this latter possibility - - they were not designed to do so, but rather to establish that D I V A constitutes the major physiological relay to HA, rather than cDLP via NI/NC. In this they were successful. The reverse is also a possibility, viz. that NI/NC receives some of its somatosensory input indirectly via the pathway from D I V A to H A to NI/NC, but the fact that the latencies of somatosensory potentials recorded in NI/NC are generally shorter than those in H A (present results, confirming those of Delius and Bennetto TM)would seem to imply that NI/NC receives its major somatosensory input directly from cDLP. Non-somatosensory inputs to H A and NI/NC Like the site of terminations of somatosensory input within the anterior hyperstriatum, I H A is also the major terminal site of visual thalamic projections to the visual Wulst 51'79, which is generally said to lie immediately caudal to the anterior somatosensory area 18'51. However, it is still not entirely clear to what extent the visual and somatosensory components of the Wulst are separate in the pigeon. Despite the numerous studies of the thalamofugal visual system (e.g. reviews, refs. 20, 25), there are few which define the precise location of thalamofugal visual terminations within the Wulst 51'76'79, and in the pigeon the complete rostrocaudal extent of these terminations is not clear. In the owl the anterior non-visual Wulst forms a small but distinct dorsomedial protuber-

218 ance 5° which is separate from the more posterior visual Wulst 51, but this is not the case in pigeon. However, the 2-deoxyglucose study of Striet et al. 75 would seem to suggest that the visual area extends rostrally into the somatosensory regions of H A as defined in the present experiments. Furthermore, the thalamic location of D I V A appears to overlap that of the origin of projections to the visual Wulst 6'61, although for the most part D I V A lies rather more medially and more caudally. But until single unit studies directly examine the possibility of polysensory input to H A neurons, the question of whether the somatosensory and visual components of H A are separate must remain open. The possibility of some overlap of somatosensory and visual areas must also be raised with respect to NI/NC. Gamlin and Cohen 2s'29 have previously defined the NI/NC projection of cDLP (a rostral component of which was also shown by Kitt and Brauth 55) and further shown that neurons within cDLP are responsive to visual stimulation, a finding which is congruent with the deep tectal projection to this nucleus 29'34. The findings of Gamlin and Cohen 29 imply, therefore, that the terminal field of cDLP within NI is the end station of a second ascending visual tectofugal pathway, and indeed, visual responses have been recorded within the caudomedial neostriatum 3°. And yet a somatosensory component of the cDLP projection field was very clear in the present study, but it did not extend further rostrally than approximately A7.50, which is at the caudal end of the cDLP projection field within NI. Again single unit studies will be required to resolve whether and to what extent there is convergence of somatosensory and visual projections within NI, but on the basis of currently available data it appears that the majority of the cDLP projection field is visual, with a smaller, caudal part being either somatosensory, or visual and somatosensory. The latter is certainly a possibility, since the deep tectal layers which give rise to the projection to c D L P 29'34 are where polysensory units can be found 17. It should also be emphasized in this context that in the present study the neurons in cDLP which were retrogradely labelled from injections in NI/NC at physiologically identified sites were predominantly in the ventral part of this nucleus, the part which receives a rather diffuse projection from the dorsal column and external cuneate nuclei s°.

To complicate matters, somatosensory and auditory overlap within both H A and NI/NC is also a possibility, particularly the latter. Auditory information could conceivably reach H A from NI/NC, which not only includes the cDLP terminal field but also parts of the Ov terminal field (i.e. Field L of Rose 73) lying more medially and dorsally. Although previous studies of the efferents of Field L in birds 12'52 have not noted a projection to HA, auditory responses have been recorded in this general region 1,a8. With regard to NI/NC itself, the diagrams of Karten 43 depicting the terminal field of Ov within the caudomedial neostriatum apparently show it to partly overlap that of the cDLP projection as defined in the present study and that of Gamlin and Cohen 29. Undoubtedly, the majority of the Ov terminal field lies caudal to that of the cDLP terminal field, but there could be a region of overlap of the two fields in the region of the C10, immediately dorsal to MLD. This could explain how Erulkar 26 was able to record somatosensory and auditory evoked potentials at the same site within the caudomedial neostriatum (confirmed by Wild, unpublished observations), and why in the present study all NI/NC injections located more ventrally, within or near CIO, labelled both cDLP and Ov neurons, whereas injections located more dorsally or caudally labelled only Ov neurons. This does not take account of the fact, however, that somatosensory evoked potentials up to 300 #V in amplitude and/or multiunit activity could be and were recorded at these more caudal sites, as they were by Delius and Bennetto is. Perhaps these regions receive a somatosensory input from sources other than caudal DLP, e.g. the anterior H A (present results), or even Ov itself (see below). Certain sites within the neostriatum studied in the present experiments, some of which were indubitably responsive to body stimulation, were also responsive to mechanical stimuli applied to the beaks and head, and to the stereotax, etc. Yet units at these sites could not be driven by electrical stimulation of the beaks or mandibular nerve, and in the absence of any trigeminal projections to the caudomedial neostriatum s2, it appears that such mechanical stimuli are transmitted by bone conduction to the auditory apparatus and hence to Field L. It may well be that these findings are simply an artifactual result of holding the head in the stereotax with ear bars, but the

219 sheer sensitivity of the unit responses in the neostriatum might also suggest a physiological role for this kind of sound transmission. Within the context of audiosomatic convergence, it is interesting to note that the caudomedial neostriatum is not the only region of such convergence in the avian telencephalon. As noted above, the major trigeminal representation of the beaks of birds is located quite separately from the representation of the rest of the body 1°'83, within the rostroventrally located NB. Within and near this nucleus, short latency auditory potentials have repeatedly been recorded 18'32'53'54 and it has recently been shown in the pigeon that NB receives a direct projection not only from the principal trigeminal nucleus 82, but also from the nucleus ventralis lemnisci, pars ventralis 4, confirming previous reports in the starling 53. It is also interesting to note that in their physiological study of NB in the pigeon, Witkovsky et al. 83 also found units which were very responsive to touching the stereotax, table, etc., in similar fashion to the findings of the present study with respect to the caudomedial neostriatum.

Comparative considerations What the functional significance of audiosomatic convergence is for the pigeon is a matter of speculation 19, but it is clear that such convergence is not confined to birds: it is even more extensive in some reptiles, e.g. turtles. The work of Belekhova and colleagues s has shown, for instance, that in these animals auditory and somatosensory projections overlap at all brain levels. At the thalamic level units responsive to auditory and somatosensory stimuli can be found within the same nucleus, namely nucleus reuniens (see Belekhova et al.S), which, like Ov of birds, receives an auditory projection from an auditory midbrain nucleus 8'27'43'68'69 but, unlike Or, also a somatosensory projection directly from the spinal cord 8'56. However, both nucleus reuniens in turtles and Ov in birds, receive a projection from the ICO of the midbrain 8 (also Wild, unpublished observations), which in various reptiles and in birds in turn receive a projection from the dorsal column nuclei and spinal cord 5'8'24'41'56. Nucleus reuniens and Ov both then project to the ventromedial part of D V R 7' 8,12,43,69. Thus, in addition to the more direct pathway for somatosensory information to reach the D V R via

cDLP in birds, or via nucleus medialis posterior in some reptiles 7'7°, there is in the pigeon a potentially indirect route via ICO and Ov (or periovoidal fields) similar to the pathway in turtles reaching the D V R via ICO and nucleus reuniens. Audiosomatic convergence within the thalamus and neocortex of mammals is also well documented 3'11'14"67, although the routing of somatosensory input through the medial geniculate complex appears to involve the medial or magnocellular nucleus, or posterior thalamic nuclei, rather than the ventral nucleus 3 with which nucleus ovoidalis of birds is thought to be homologous 43. Also, within the parietal neocortex, auditory and somatosensory areas seem to be more clearly separated than in birds and reptiles except for a transitional zone between SII and the auditory fields TM. It is particularly interesting to note, however, that in the grey squirrel a parietal area ventral to SII has been discovered 57, within which auditory and somatosensory inputs extensively overlap. The presence of two reciprocally connected somatosensory body areas in the telencephalon of the pigeon inevitably invites a comparison with the neocortical organization of SI and SII in mammals, although it is recognized that the evolution of such characters may well be of a convergent homoplastic rather than an homologous kind 65. In the pigeon, the rostrally located area, HA, receives its thalamic input from D I V A , which appears to be the major diencephalic recipient of medial lemniscal fibers 8°'85. In this respect D I V A would be analogous to the VPL or VB nucleus of mammals, and H A to SI. Clearly, however, there are important differences and many unknowns. For instance, one difference between H A and SI is that both sides of the body are represented in HA, a feature not present in SI, for VPL receives a purely contralateral projection, whereas D I V A receives a bilateral projection, albeit with a contralateral predominance s°. Unknowns are the degree to which H A in birds is somatotopically organized, or contains an ordered representation of somatosensory submodalities. The present and other 18 data in pigeon, showing that many units in H A have wide receptive fields and that some can even be driven by radial and sciatic nerve stimulation, suggest that, unlike SI, H A is not strictly somatotopically organized; but completely systematic, fine-detailed microelectrode mapping studies have yet to be carried out, and some findings

220 of the present study and that of a discrete 'toe area' in the anterior Wulst of the owl 5° are suggestive of some kind of ordered body representation. The second of the two somatosensory areas in the pigeon, lying within NI/NC also shows certain similarities to, and differences from, SII of mammals. Like SII, it lies behind SI (although in the pigeon quite far removed), is closely related to the auditory projection fields, and contains a bilateral representation of the body. Unlike SII, however, it appears to receive its somatosensory input from the thalamus, at least in large part, from a different nucleus (viz. caudal DLP) than that which projects to the primary somatosensory area (viz. DIVA). In the present experiments the projection of cDLP was found to be specific to NI/NC and did not include H A (cf. refs. 6, 29). In contrast, in mammals both SI and SII receive their input from VP, in part from the same neurons TM. However, regions immediately caudal to SII (e.g. in the more posterior parts of the cat's anterior ectosylvian gyrus or primate retroinsular area), areas which are transitional with auditory fields, are part of an extensive projection zone of the posterior thalamic nuclear complex 33, different nuclei of which receive differential ascending projections from the superior and inferior colliculi, spinothalamic tract and perhaps dorsal column nuclei 35. Somewhat analogously in the pigeon, cDLP also receives a projection from deep layers of the tectum 29'34, possibly the spinal cord 41, and dorsal column nuclei 5'8°. It does not appear to receive a projection from the inferior colliculus (MLD), however. Nevertheless, as a working hypothesis, cDLP could perhaps be regarded as a 'Potype' nucleus, and NI/NC to which it and Ov project, some rostrocaudally ordered composite of a visual area, an SII-type area, a transitional somatosensoryauditory area, and a purely auditory area. With respect to the output of the two somatosenso-

ry areas delineated in this and other studies, it should be noted that HA, which includes the anterior somatosensory area and which receives a projection from the caudal somatosensory area and perhaps parts of the auditory Field L, is also the origin of the septomesencephalic tract (TSM) 71. The descending projections of the basal branch of this tract 21'4°'77'78 have been likened to those of the mammalian pyramidal tract 45'5a'84. As such, TSM represents an important route by which motor commands, set up by somatosensory, visual, and perhaps auditory input, can reach lower centers; a route separate and different from the OM - - comparable with a variant of the mammalian pyramidal tract 3t - - the origin of which in the archistriatum s6 is also subject to the indirect influences of somatosensory (beak) 82, visual 72, and auditory 67 input. In the present experiments W G A - H R P injections into the anterior H A anterogradely labelled the lateral aspect of the basal branch of TSM which terminated for the most part in the nucleus spiriform medialis (SpM) of the caudal thalamus. This nucleus is known to provide the cerebellum with a massive mossy fiber inpu( 5'49 and also projects to the precerebellar medial pontine nucleus (Wild, unpublished observations). Since it receives a projection also from OM 86, SpM seems to be an important relay, in addition to the pontine nuclei themselves ~5"51'86, through which sensory information processed at higher levels reaches the cerebellum.

ABBREVIATIONS

DLL DLM cDLP rDLP DMA DMP DVR E FPL G

BO

C Cb CE CIO DIP DIVA DLA

Bulbus olfactorius Nucleus cuneatus Cerebellum Nucleus cuneatus externus Capsula internus occipitalis Nucleus dorsointermedius posterior Nucleus dorsalis intermedius ventralis anterior Nucleus dorsolateralis anterior

ACKNOWLEDGEMENTS This work was supported by the New Zealand Lottery Board, the Medical Research Council of New Zealand, and the New Zealand University Grants Committee. The histological assistance of Mrs. S. Braan-Stroo is gratefully acknowledged.

Nucleus dorsolateralis anterior, pars lateralis Nucleus dorsolateralis anterior, pars medialis Nucleus dorsolateralis posterior, pars caudalis Nucleus dorsolateralis posterior, pars rostralis Nucleus dorsomedialis anterior Nucleus dorsomedialis posterior Dorsal ventricular ridge Ectrostriatum Fasciculus prosencephali lateralis Nucleus gracilis

221 GG HA HD HIP HIS HV ICO IHA LMD N NB NC NI OM Ov PA

Gasserian ganglion Hyperstriatum accessorium Hyperstriatum dorsale Tractus habenulo-interpeduncularis Hyperstriatum intercalatus superior Hyperstriatum ventrale Nucleus intercoUicularis Intercalated hyperstriatum accessorium Lamina medullaris dorsalis Neostriatum Nucleus basalis Neostriatum caudale Neostriatum intermedium Tractus occipitimesencephalicus Nucleus ovoidalis Paleostriatum augmentatum

REFERENCES 1 Adamo, N.J. and King, R.L., Evoked responses in the chicken telencephalon to auditory, visual, and tactile stimulation, Exp. Neurol., 17 (1967) 498-504. 2 Adams, J.L., Stabilizing and rapid thionin staining of TMBbased HRP reaction product, Neurosci. Lett., 17 (1980) 7-9. 3 Aitken, L.M., Irvine, D.R.F. and Webster, W.R., Central neural mechanisms of hearing. In I. Darian-Smith (Ed.),

Handbook of Physiology, Section I: The Nervous System; Vol. IlL Sensory Processes, Part 2, American Physiological Society, Bethesda, Maryland, 1984. 4 Arends, J.J.A. and Zeigler, H.P., Anatomical identification of an auditory pathway from a nucleus of the lateral lemniscus to the frontal telencephalon (nucleus basalis) of the pigeon, Brain Research, 398 (1986) 375-381. 5 Arends, J.J.A., Woelders-Blok, A. and Dubbeldam, J.L., Efferent connections of the nuclei of the descending trigeminal tract in the mallard (Anus platyrhyncos L.), Neuroscience, 13 (1984) 797-817. 6 Bagnoli, P. and Burkhalter, A., Organization of the afferent projections to the Wulst in the pigeon, J. Comp. Neurol., 214 (1983) 103-113. 7 Balaban, C.D. and Ulinski, P.S., Organization of thalamic afferents to anterior dorsal ventricular ridge in turtles. I. Projections of thalamic nuclei, J. Comp. Neurol., 200 (1981) 95-129. 8 Belekhova, M.G., Zharskaja, V.D., Khachunts, A.S., Gaidaenko, G.V. and Tumanova, N.L., Connections of the mesencephalic, thalamic and telencephalic auditory centers in turtles. Some structural bases for audiosomatic interrelations, J. Hirnforsch., 26 (1985) 127-152. 9 Benowitz, L., Functional organization of the avian telencephalon. In S.O.E. Ebbesson (Ed.), Comparative Neurology of the Telencephalon, Plenum, New York, 1980, pp. 389-421. 10 Berkhoudt, H., Dubbeldam, J.L. and Zeilstra, S., Studies on the somatotopy of the trigeminal system in the mallard, Anus platyrhyncos L. IV. Tactile representation in the nucleus basalis, J. Comp. Neurol., 196 (1981) 407-420. 11 Berman, A.L., Overlap of somatic and auditory cortical response fields in anterior ectosylvian gyrus of cat, J. Neurophysiol., 24 (1961) 595-607.

PP PrV RSd Rt SG SPC SpM SRt ST T TIO TO TOv TSM TI'D V

Paleostriatum primitivum Nucleus sensorius principalis nervi trigemini Nucleus reticularis superior, pars dorsalis Nucleus rotundus Substantia gelatinosa Nucleus superficialis parvocellularis Nucleus spiriformis medialis Nucleus subrotundus Nucleus subtrigeminalis Nucleus triangularis Tractus isthmo-opticus Tractus opticus Tractus ovoidalis Septomesencephalic tract Nucleus et tractus descendens nervi trigemini Ventriculus

12 Bonke, B.A., Bonke, D. and Scheich, H., Connectivity of the auditory forebrain nuclei in the Guinea Fowl (Numida meleagris), Cell Tissue Res., 200 (1979) 101-121. 13 Burton, H. and Robinson, C.J., Organization of the SII parietal cortex. In C.N. Woolsey (Ed.), Cortical Sensory Organization, Vol. III. Multiple Somatic Areas, Humana Press, Clifton, New Jersey, 1981, pp. 67-119. 14 Carreras, M. and Andersson, S.A., Functional properties of neurons of the anterior ectosylvian gyrus of the cat, J. Neurophysiol., 26 (1963) 100-126. 15 Clarke, P.G.H., Some visual and other connections of the cerebellum of the pigeon, J. Comp. Neurol., 174 (1977) 535-552. 16 Cohen, D.H. and Karten, H.J., The structural organization of avian brain: an overview. In I.J. Goodman and M.W. Schein (Eds.), Birds: Brain and Behavior, Academic, New York, 1974, pp. 29-73. 17 Cotter, J.R., Visual and non-visual units recorded from the optic tectum of Gallus domesticus, Brain, Behav. Evol., 13 (1976) 1-21. 18 Delius, J.D. and Bennetto, K., Cutaneous sensory projections to the avian forebrain, Brain Research, 37 (1972) 205-221. 19 Delius, J.D., Runge, T.E. and Oeckinghaus, H., Short-latency auditory projection to the frontal telencephalon of the pigeon, Exp. Neurol., 63 (1979) 594-609. 20 Donovan, W.J., Structure and function of the pigeon visual system, Physiol. Psychol., 6 (1978) 403-437. 21 Dubbeldam, J.L., The basal branch of the septo-mesencephalic tract in the mallard, Anus platyrhyncos L., Acta. Morphol. Neerl. Scand., 14 (1976) 98. 22 Dubbeldam, J.L. and Karten, H.J., The trigeminal system in the pigeon (Columba livia). I. Projections of the gasserian ganglion, J. Comp. Neurol., 180 (1978) 661-678. 23 Dubbeldam, J.L., Brauch, C.S.M. and Don, A., Studies on the somatotopy of the mallard, Anas platyrhynos L. III. Afferents and organization of nucleus basalis, J. Comp. Neurol., 195 (1981) 391-405. 24 Ebbesson, S.O.E., Brainstem afferents from the spinal cord in a sample of reptilian and amphibian species, Ann. N.Y. Acad. Sci., 167 (1969) 80-102. 25 Emmerton, J., Functional morphology of the visual system. In M. Abs (Ed.), Physiology and Behaviour of the Pigeon, Academic, London, 1983, pp. 221-244.

222 26 Erulkar, S.D., Tactile and auditory areas in the brain of the pigeon. An experimental study by means of evoked potentials, J. Comp. NeuroL, 103 (1955) 421-458. 27 Foster, R.E. and Hall, W.C., The organization of central auditory pathways in a reptile, Iguana iguana, J. Comp. Neurol., 178 (1978) 783-831. 28 Gamlin, P.D.R. and Cohen, D.H., A possible second ascending avian tectofugal pathway, Soc. Neurosci. Abstr., 8 (1982) 206. 29 Gamlin, P.D.R. and Cohen, D.H., A second ascending visual pathway from the optic tectum to the telencephalon in the pigeon (Columba livia), J. Comp. Neurol., 250 (1986) 296-310. 30 G~intfirkiin, O., Evidence for a third primary visual area in the telencephalon of the pigeon, Brain Research, 294 (1984) 247-254. 31 Haartsen, A.B. and Verhaart, W.J.C., Cortical projections to the brainstem and spinal cord of the goat by way of the pyramidal tract and the bundle of Bagley, J. Comp. Neurol., 129 (1967) 184-202. 32 Harman, A.L. and Phillips, R.E., Responses in the avian midbrain, thalamus and forebrain evoked by click stimuli, Exp. Neurol., 18 (1967) 276-286. 33 Heath, C.J. and Jones, E.G., An experimental study of ascending connections from the posterior group of thalamic nuclei in the cat, J. Comp. Neurol., 141 (1971) 397-426. 34 Hunt, S.P. and Ktinzle, H., Observations on the projections and intrinsic organization of the pigeon optic rectum: an autoradiographic study based on anterograde and retrograde axonal and dendritic flow, J. Comp. Neurol., 170 (1976) 153-172. 35 Jones, E.G., The posterior complex of nuclei. In E.G. Jones, The Thalamus, Plenum, New York, 1985, pp. 573-604. 36 Kaas, J.H., The segregation of function in the nervous system: why do sensory systems have so many subdivisions? In W.D. Neff (Ed.), Contributions to Sensory Physiology, Vol. 7, Academic, New York, 1982, pp. 201-240. 37 Kaas, J.H., What, if anything, is SI? Organization of first somatosensory area of cortex, Physiol. Rev., 63 (1983) 206-231. 38 Kaas, J.H., Nelson, R.J., Sur, M. and Merzenich, M.M., Organization of somatosensory cortex in primates. In F.O. Schmitt, F.C. Warden, G. Adelman and S.G. Dennis (Eds.), The Organization of Cerebral Cortex, MIT, Cambridge, 1981, pp. 237-261. 39 Kaas, J.H., Sur, M., Nelson, R.J. and Merzenich, M.M., The postcentral somatosensory cortex. Multiple representations of the body in primates. In C.N. Woolsey (Ed.),

Cortical Sensory Organization, Vol. I. Multiple Somatic Areas, Humana Press, Clifton, New Jersey, 1981, pp. 29-45. 40 Kalischer, O., Das Grosshirn der Papageien in anatomischer und physiologischer Beziehung, Abh. Preuss. Akad. Wiss., 4 (1905) 1-105. 41 Karten, H.J., Ascending pathways from the spinal cord in the pigeon (Columba livia), Proc. 16th Int. Zool. Cong., Part 2 (1963) 23. 42 Karten, H.J., The organization of the ascending auditory pathway in the pigeon (Columba livia), I. Diencephalic projections of the inferior colliculus (nucleus mesencephali lateralis, pars dorsalis), Brain Research, 6 (1967) 409-427. 43 Karten, H.J., The ascending auditory pathway in the pigeon (Columba livia), II. Telencephalic projections of the

nucleus ovoidalis thalami, Brain Research, 11 (1968) 134-153. 44 Karten, H.J., The organization of the avian telencephalon and some speculations on the phylogeny of the amniote telencephalon, Ann. N.Y. Acad. Sci., 167 (1969) 164-180. 45 Karten, H.J., Efferent projections of the Wulst of the owl, Anat. Rec., 169 (1971) 353. 46 Karten, H.J., Visual lemniscal pathways in birds. In A.M. Granda and J.H. Maxwell (Eds.), Neural Mechanisms of Behavior in the Pigeon, Plenum, New York, 1979, pp. 409-430. 47 Karten, H.J. and Revzin, A.M., The efferent connections of the nucleus rotundus in the pigeon, Brain Research, 2 (1966) 368-377. 48 Karten, H.J. and Hodos, W., A Stereotaxic Atlas of the Brain of the Pigeon (Columba livia), Baltimore, Johns Hopkins University Press, 1967. 49 Karten, H.J. and Finger, T.E., A direct thalamo-cerebellar pathway in pigeon and catfish, Brain Research, 102 (1976) 335-338. 50 Karten, H.J., Konishi, M. and Pettigrew, J.D., Somatosensory representation in the anterior Wulst of the owl (Speotyto cunicularis), Soc. Neurosci. Abstr., 4 (1978) 554. 51 Karten, H.J., Hodos, W., Nauta, W.J.H. and Revzin, A.M., Neural connections of the 'Visual Wulst' of the avian telencephalon. Experimental studies in the pigeon (Columba livia) and owl (Speotyto cunucularia), J. Comp. Neurol., 150 (1973) 253-278. 52 Kelley, D.B. and Nottebohm, F., Projections of a telencephalic auditory nucleus - - Field L - - in the canary, J. Comp. Neurol., 183 (1979) 455-470. 53 Kirsch, M., Antworteigenschaften und Verbindungen des akustischen Zentrums mit kurzer Latenz im frontalen Vorderhirn des Staren (Sturnus vulgaris), Verh. Dtsch. Zool. Ges., (1983) 296. 54 Kirsch, M., Coles, R.B. and Leppelsack, H.-J., Unit recordings from a new auditory area in the frontal neostriatum of the awake starling (Sturnus vulgaris), Exp. Brain Res., 38 (1980) 375-380. 55 Kitt, C.A. and Braflth, S.E., A paleostriatal-thalamic-telencephalic path in pigeons, Neuroscience, 7 (1982) 2735-2751. 56 Kunzle, H. and Woodson, W., Mesodiencephalic and other target regions of ascending spinal projections in the turtle, Pseudemys scripta elegans, J. Comp. Neurol., 212 (1982) 349-364. 57 Luethke, L.E., Krubitzer, L. and Kaas, J.H., Connections of auditory cortex in squirrels, Soc. Neurosci. Abstr., 11 (1985) 33. 58 Merzenich, M.M. and Kaas, J.H., Principles of organization of sensory-perceptual systems in mammals. In J.M. Sprague and A.N. Epstein (Eds.), Progr. Psychobiol. Physiol. Psychol., Vol. 9, Academic, New York, 1980, pp. 1-42. 59 Merzenich, M.M., Sur, M., Nelson, R.J. and Kaas, J.H., Organization of the SI cortex. In C.N. Woolsey (Ed.), Cor-

tical Sensory Organization, Vol. 1. Multiple Somatic Areas, Humana, Clifton, New Jersey, 1981, pp. 47-66. 60 Mesulam, M.-M., Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents, J. Histochem., 26 (1978) 106-117. 61 Miceli, D. and Reperant, J., Thalamo-hyperstriatal projec-

223

62

63

64

65

66

67

68

69

70

71

72

tions in the pigeon (Columba livia) as demonstrated by retrograde double-labeling with fluorescent tracers, Brain Research, 245 (1982) 365-371. Miceli, D. and Reperant, J., Telencephalic afferent projections from the diencephalon and brainstem in the pigeon. A retrograde multiple-label fluorescent study, Exp. Biol., 44 (1985) 71-99. Nauta, W.J.H. and Karten, H.J., A general profile of the vertebrate brain, with sidelights on the ancestry of cerebral cortex. In G.C. Quarton, T. Melnechuk and F.O. Scbmitt (Eds.), The Neurosciences: Second Study Program, Rockefeller Press, New York, 1970, pp. 7-26. Necker, R., Somatosensory system. In M. Abs (Ed.), Physiology and Behaviour of the Pigeon, Academic, London, 1983, pp. 169-192. Northcutt, R.G., Evolution of the telencephalon in nonmammals. In W.M. Cowan, Z.W. Hall and E.R. Kandel (Eds.), Ann. Rev. Neurosci., 4, 1981, pp. 301-350. Nottebohm, F., Kelley, D.B. and Paton, J.A., Connections of vocal control nuclei in the canary telencephalon, J. Comp. Neurol., 207 (1982) 344-357. Poggio, G.F. and Mountcastle, V.B., A study of the functional contributions of the lemniscal and spinothalamic systems to somatic sensibility, Bull. Johns Hopkins Hosp., 106 (1960) 266-316. Pritz, M.B., Ascending connections of a midbrain auditory area in a crocodile, Caiman crocodilus, J. Comp. Neurol., 153 (1974) 179-198. Pritz, M.B., Ascending connections of a thalamic auditory area in a crocodile, Caiman crocodilus, J. Comp. Neurol., 153 (1974) 199-214. Pritz, M.B. and Northcutt, R.G., Anatomical evidence for an ascending somatosensory pathway to the telencephalon in crocodiles, Caiman crocodilus, Exp. Brain Res., 40 (1980) 342-345. Reiner, A. and Karten, H.J., The laminar source of efferent projections from the avian Wulst, Brain Research, 275 (1983) 349-354. Ritchie, T., lntratelencephalic Visual Connections and

Their Relationships to the Archistriatum in the Pigeon (Columba livia), Unpublishes doctoral dissertation, University of Virginia, Charlottesville, 1979. 73 Rose, M., Ober die cytoarchitektonische Gliederung des

Vorderhirns der Vogel, Zeit. Psychol. Neurol., 21 (1914) 278-352. 74 Spreafico, R., Hayes, N.L. and Rustioni, A., Thalamic projections to the primary and secondary somatosensory cortices in cat: single and double retrograde tracer studies, ]. Comp. Neurol., 203 (1981) 67-90. 75 Streit, P., Burkhalter, A., Stella, M. and Cuenod, M., Patterns of activity in pigeon's brains visual relays as revealed by the [14C]2-deoxyglucose method, Neuroscience, 5 (1980) 1053-1066. 76 Streit, P., Stella, M. and Cuenodl M., Transneuronal labeling in the pigeon visual system, Neuroscience, 5 (1980) 763775. 77 Verhaart, W.J.C., Forebrain bundles and fibre systems in the avian brainstem, J. Hirnforsch., 13 (1971) 39-64. 78 Wallenberg, A., Die basalen Aste des Scheidewandbundels der Vogel, Anat. Anz. 28 (1906) 394-400. 79 Watanabe, M., Ito, H. and Masai, H., Cytoarchitecture and visual receptive neurons in the Wulst of the Japanese quail (Coturnix coturnix japonica), J. Comp. Neurol., 213 (1983) 188-198. 80 Wild, J.M., Connectional anatomy of the somatosensory system of the pigeon, Soc. Neurosci. Abstr., 9 (1983) 244. 81 Wild, J.M., The avian somatosensory system. I. Primary spinal afferent input to the spinal cord and brainstem in the pigeon (Columba livia), J. Comp. Neurol., 240 (1985) 377-395. 82 Wild, J.M., Arends, J.J.A. and Zeigler, H.P., Telencephalic connections of the trigeminal system in the pigeon (Columba livia): a trigeminal sensorimotor circuit, J. Comp. Neurol., 234 (1985) 441-464. 83 Witkovsky, P., Zeigler, H.P. and Silver, R., The nucleus basalis of the pigeon: a single unit analysis, J. Comp. Neurol., 147 (1973) 119-128. 84 Zecha, A., The 'pyramidal tract' and other telencephalic efferents in birds, Acta Morphol. Neerl. Scand., 4 (1964) 194-195. 85 Zecha, A., The medial lemniscus in the pigeon, Acta Morphol. Neerl. Scand., 6 (1966) 322. 86 Zeier, H. and Karten, H.J., The archistriatum of the pigeon: organization of afferent and efferent connections, Brain Research, 31 (1971) 313-326.